Capture wildlife or unsuspecting visitors on your DSLR with this PIR-based camera trigger.

BUILD TIME: 1.5 HOURS + 3D PRINTING TIME
DIFFICULTY RATING: Intermediate

If you’re a photography enthusiast, you are no doubt aware of how difficult it can be to take the perfect natural looking photo of wildlife. This is largely due to wildlife being particularly cautious of humans, generally for good reasons. This gives photographers a couple of options; they can attempt to hide and patiently wait for their apprehensive and timid subject or they can set up what is known as a camera trap. This trap is essentially an autonomous DSLR camera with a sensor built in to detect the subject. When the sensor detects a subject, the camera takes a photo. This means the photographer does not need to be anywhere near the camera and the subject, which means that the animal can be photographed in their most natural state possible.

THE BROAD OVERVIEW

We present you with two methods to trigger your DSLR when movement is detected via the PIR sensor.

First, we present you with a microcontroller method that uses an Arduino Nano to trigger the camera, providing the necessary delays and signals for effective single or multiple shots. This method also enables you to add your own custom settings to suit your application.

The second method uses the circuitry on the PIR sensor itself, instead of using a microcontroller. Integrated potentiometers on the PIR circuit enable us to control some of the trigger settings we need for effective shots.

We include a rechargeable Lithium battery feature, which allows the project to operate in the outdoors away from mains power, and it is all designed in a compact 3D printed housing that can conveniently fit onto your camera’s hot shoe.

The design is tested on various Canon models with a 2.5mm stereo plug input, but can be adapted to other camera brands quite easily.

HOW IT WORKS

PIR Sensor

The heart of this project is the Passive Infrared (PIR) sensor module which is commonly available from electronics retailers. This module is, in effect, a pyroelectric sensor that generates a very tiny voltage when exposed to tiny variations in heat. We know what you’re thinking “How does detecting small changes in temperature give us any idea if an animal is moving?” Therefore, its likely best if we explain the entire process.

Every living creature on Earth generates heat as the byproduct of turning their food into energy, and using that energy to do work. If the creature is large enough, we can measure the heat being produced.

This heat is infrared radiation, which is on a wavelength a little longer than visible light, meaning we humans are incapable (for the most part) of being able to physically view this radiation. However, crystals have the ability to become polarised i.e. generate small electrical charges on their surface when experiencing a temperature differential.

To generate this temperature differential, the PIR sensor uses a clever optical device called a Fresnel lens. This is the dome you can see on the outside of the sensor itself. This lens is made up of many individual lenses that direct the incoming infrared radiation onto the sensor, which consists of two series connected crystal elements. When something generating heat and with sufficient mass moves in front of the Fresnel lens, the lens directs that infrared energy towards the sensor. As the object/animal moves past the sensor, the energy level of the infrared hitting the sensor also changes. This change generates a very tiny voltage that is then amplified so that the signal can be measured easily.

In these inexpensive sensors, this signal is fed into a specialised integrated circuit designed specifically for PIR sensors called the BISS0001. This IC handles all of the amplifying and comparator operations including the delay time and sensitivity of the sensor. This means the module output is simply a high voltage when movement is detected or a low voltage when no movement is detected. Naturally, this makes this sensor incredibly easy to use.

Camera Trigger

Unfortunately, there isn’t a standard among DSLR camera manufactures for their remote triggering methods. As such, it is impossible to create a single solution that is compatible with all cameras. However, most cameras follow a similar trigger method, albeit done with different connectors, etc.

Our testing was done using a Canon 80D and the 1200D, which use a 2.5mm stereo plug, commonly used for audio applications. Your camera may have a different connector, which you will need to allow for if you are building this project. If you are using a Canon DSLR, you can check if this project will work with your DSLR by searching for the official Canon remote shutter RS-60E3. If your camera is compatible with this remote shutter, this project should work for you.

In the case of the Canon 80D, the tip and ring connection on the plug shown here are held to 3.3V via a pullup resistor, and the sleeve is tied to the cameras ground/common. The camera has a physical switch with two positions; the first position allows the camera to automatically focus (AF) the shot based on your AF settings, and the second position signals the camera to actuate the shutter mechanism.

The switch itself makes a physical connection between the camera’s ground and the pullup resistor for the shutter, and focus signal wires causing the controller to see a 0V potential at these pins, after which the controller reacts by initiating the focus sequence or actuating the shutter.

This means it is trivial to tell the camera to actuate the focus and shutter processes, as all we need to do is pull either and/or both of the connections down to the camera’s ground potential. This can be done using a microcontroller or even the PIR sensor itself.

This is our interpretation of the camera circuit, based on tests using a multimeter on our Canon DSLR. The switch used in the camera is likely a DPDT switch actuating both signals, however, we have shown it this way for simplicity.

OPTOCOUPLER

DSLR cameras are expensive, so we are using an optocoupler to galvanically isolate our circuit from the camera’s internal circuitry. This will help ensure that the camera and our circuit do not share common ground, and thus any voltage potential in our circuitry is 100% isolated from the camera. i.e. current cannot flow into the camera from the project.

This isolation is made simple thanks to the Fairchild’s 4N25 optocoupler. These devices consist of a gallium arsenide IR LED pointed at a phototransistor.

When current from our circuit flows into pin 1 of the IC package, it illuminates an internal IR LED, with current flowing out of the package via pin 2. This infrared light causes the base of the internal phototransistor to saturate, which allows current to flow into pin 5 of the device and out of pin 4.

This means that the camera circuit is isolated from the detector circuit, having no electrical connection between the two. Therefore, all our circuitry needs to do is illuminate the internal LED in this optocoupler, which in turn will allow us to trigger the camera remote trigger functions.

The Prototype:

We got started on this project thinking we needed to use a microcontroller in order to control the camera, and as such, our prototype is quite a bit different from the final project.

We built the prototype slightly different from the Fritzing diagram shown as we were largely experimenting with the circuit and code. Thus, we added a switch that would replace the sensor and allow us to simulate a sensor trigger at will. This excluded potential sensor issues during the testing and design phase. We also added indicator LEDs so we could physically see when the optocouplers were triggered and therefore triggering the camera functions. These extra elements are excluded from the final Fritzing diagram for simplicity.

* Quantity required, may only be sold in packs. Breadboard and prototyping hardware also required.

ELECTRONICS

The electronics are fairly self-explanatory for this project. The only potential issue is the fact that the optocoupler is polarity dependent. This means you need to be sure that the camera ground is connected to pin 4 of the 4N25 optocoupler and the shutter or focus signal from the camera is connected to pin 5. If you get these around the wrong way current will not flow through the phototransistor and thus the camera will not respond. Rest assured though, getting it around the wrong way will not cause any damage to the device, it will simply not function.

We opted to use 100Ω resistors for the optocouplers, as the datasheet for the 4N25 stated a forward voltage (Vf) of 1.5V and a maximum forward current (If) of 60mA. To be safe, we halved the maximum forward current to give us the desired current through the LED of 30mA.

R = (Vin - Vf) / If = (5 - 1.5) / 0.03 = 116.66Ω

We just rounded this to the nearest resistor value we had in our collection, which was 100Ω. This allows 35mA to flow through the LED, which is perfectly acceptable. If you don’t have a 100Ω a 120Ω would suffice too.

You can find the datasheet here: http://www.vishay.com/docs/83725/4n25.pdf

THE CODE

The code can be downloaded from our website. You will notice it is fairly simple. We have added customisable variables that easily allow you to make changes to suit your specific needs and requirements.

Focus Time adjusts the number of milliseconds between triggering the focus function and then taking the picture. This delay is a must, as it can take the camera a little while depending on the lens used to find focus. In our tests, we used a Canon 50mm prime lens which had a very quick autofocus and, as such, we found that 500ms (half a second) was sufficient. You may find that this is too short for your lens, so increasing this value will allow for that.

Min Time sets the delay after triggering, and is used to prevent the same trigger event causing multiple unwanted camera triggers. It is simply a delay of one second in our case. This means after taking the picture or sequence of pictures, the camera will wait for 1 second before resuming to poll the sensor pin for a high input. During this time, the camera will not take more pictures.

Shots sets the number of photos that will be taken after a trigger event. It is used to force the camera to repeatedly take x number of images after a single trigger event, irrespective of the state of the sensor pin. In our instance, we take 5 shots with each one re-focussing before taking the shot.

Pause is used to set the time between shots and gives the camera sufficient time to write to the memory card, and to detect that the shutter pin has changed states, which from our tests seems to need about 10 to 20ms. However, this will depend on the camera and the memory card you’re using.

SETUP

int shutter = 4;
int focus = 3;
int sensor = 2;
int focusTime = 500; 
// delay to allow the camera to auto focus
int minTime = 1000; 
// minimum time before a sensor 
// re-trigger can initiate a sequence
int shots = 5; 
// number of photos taken per trigger
int pause = 50; 
// mili secs between camera taking shots
void setup() {
  pinMode(shutter, OUTPUT);
  pinMode(focus, OUTPUT);
  pinMode(sensor, INPUT);
} 

In summary, the code simply polls the sensor pin, constantly checking if it has gone high. If it detects the sensor has been trigged the code jumps into a For loop that is repeated the number of times stored in the shots variable. In each loop, the program pulls the cameras focus line to the cameras ground potential, waits the number of milliseconds stored in the focus time variable, before also pulling the cameras shutter line to the camera’s ground potential, and then pausing for the number of milliseconds stored in the pause variable.

After this, the camera shutter and focus lines are allowed to be pulled back to their 3.3V state, and a small delay of 20ms gives the camera time to detect this change of state.

Once this has been repeated as many times as specified in the shots variable, the program goes into another delay to prevent extra re-triggering. This delay is set using the min time variable and by default is 1000 milliseconds. After this delay, the program simply goes back to polling the sensor pin.

TESTING THE PROTOTYPE

To test the prototype, we recommend that you attach a pushbutton switch to the circuit to simulate the sensor actuating. This way, you can “on-demand” test the device without worrying about triggering the sensor. This is because it’s difficult to get consistent results while the sensor is not fixed. We also recommend you add a couple of LEDs so you can see when the optocoupler is triggered.

With your camera connected, press the button or trigger the sensor. Your camera should focus and then take a picture. If you look at the photo taken and the result is not in focus, there can be a couple of issues.

LOOP

void loop() {
  if (digitalRead(sensor) == HIGH) {
    for (int i = 0; i < shots; i++) {
      digitalWrite(focus, HIGH);
      delay(focusTime);
      digitalWrite(shutter, HIGH);
      delay(pause);
      digitalWrite(focus, LOW);
      digitalWrite(shutter, LOW);
      delay(20);
      }
      delay(minTime);
  }
  else {
      digitalWrite(focus, LOW);
      digitalWrite(shutter, LOW);
    }
} 

1. Make sure your camera is set to use as many of the AF points as possible
2. Extend the focus time variable to give the camera more time to respond

If the camera takes more than one loop of photos with the one trigger event, you can use the potentiometers on the sensor to decrease the duration that the sensor stays high. You could also increase the min time variable.

If the camera display reads ‘busy’, ‘please wait’ or a similar error, the device may be trying to take a picture quicker than the camera can write the data. You can try to increase the pause variable to give the camera more time to store the captured image.

If the camera takes fewer photos than the number designated in the shots variable there may be an issue with the pause variable. This can easily happen in low light conditions because the shutter of the camera needs to stay open longer. This results in the camera ignoring requests to take another shot while the shutter is already open taking the first shot.

If the camera does not react at all, make sure that you have the optocoupler and 2.5mm plug wired correctly. Remember the optocoupler is not a switch but rather it is a transistor. As such, current can only flow from the collector to the emitter. If this is around the wrong way the circuit will not work.

The Main Build:

Our original plan for this design was to replace the Arduino Nano used in the prototype for an ATtiny85 microcontroller, with the aim to reduce the cost and the size of the project. However, after experimenting with the circuit, it became apparent that all of the functions we were doing with the microcontroller can be done inside the camera itself or by using the PIR sensor potentiometers. This makes the added complexity of two optocouplers and a microcontroller essentially redundant.

We can easily set the number of shots to be taken with each press of the button by using the camera’s continuous shooting modes, and setting the time potentiometer on the sensor itself to dictate how long the shutter is held down for, thus how many shots are taken.

Likewise, the focus triggering can be nullified by using the camera’s built-in continuous focus modes, called servo AI on Canon cameras or by simply manually focusing the lens. You could even tie the focus permanently to the camera’s ground to keep the focus active. Therefore, we decided that for the final design we would simplify the project and leave out the microcontroller and focus optocoupler.

* Quantity required, may only be sold in packs. The fasteners will need to be purchased from a hardware store or similar fastener retailer.

This way, we have still shown both concepts and thus the reader can decide for themselves which of the circuits best suits their needs.

Note, we have tried to make the enclosure big enough to fit either design, but it may be a little tight when using the design with two optocouplers and the ATtiny85. We recommend trial fitting the components before soldering them completely so you can make changes to the layout if the room is too tight.

With regard to power, we had to make a late change to this project design. Initially, in an attempt to conserve battery power, we completely removed the 3.3V linear voltage regulator and reverse polarity protection diode from the PIR sensor. We figured this would allow us to conserve power and supply the device directly from the 3.2 – 4.2V lithium cell. We made this decision based on the info in the datasheet for the pyroelectric sensor and the BIS0001 IC, which both stated the device will work fine with input voltages as low as 2.5V and well over 4V. Whilst, in theory, this would work perfectly fine as we were not exceeding any of the absolute maximum parameters, in practice we observed very inconsistent and intermittent results once the voltage dropped from the fully charged 4.2 to around 3.5V in this configuration.

We suspect that this erratic behavior was caused by the circuit being specifically designed to run at the stable 3.3V output from the regulator, however, without a schematic for the module or reverse engineering it, we have no idea what we would need to change for it to work with a fluctuating input voltage.

Therefore, we were faced with only one option, to increase the input voltage to 5V to accommodate for the PIR sensor’s expected input. The simplest way to do this, using retail store-bought components, was to use a boost converter to increase the varying voltage from the lithium cell to a known voltage.

We used a commonly available USB 5V boost converter specifically designed to boost a lithium cell to 5V. These are efficient switchmode voltage regulators that are designed to be used in USB battery banks. To learn how they work, be sure to check out Bob Harper’s two-part series on switchmode supplies from Issue #011 and #012 where he details the principle of operation and the design of these clever modules.

Boost Module

To keep the project compact, we decided to remove the large USB connector from the boost module. This connector isn’t needed in our project and has the potential to physically interfere with the other components of the circuit.

Note: We found it easier to remove the connector by cutting its legs, then desolder the remaining part of the leg on the board using desolder braid. You may need a high power iron or use a high temperature to remove the connector’s ground plane locating lugs. We found that ~450°C gave us a sufficient amount of heat to quickly reflow the solder on the mounting lugs. While the solder was molten on one of the lugs, we used a flat head screwdriver to gently lift up the connector to allow sufficient room to get our flush-cut side cutters in and cut the tag. Repeat to remove the lug from the opposite side.

To remove the remaining four legs from the board, we first added a fair amount of solder to the pads to reflow the lead-free solder used on these modules. This helps to give the desolder braid/solder wick enough solder to start the wicking process. After tinning the pads, you simply lay the braid over the pad/s you wish to remove the solder from and place your soldering iron tip on top of the braid. This will heat the braid and the pad to the melting point of solder. The solder on the PCB pad will wick from the pad onto the fine strands of the desolder braid. Lift up while the solder is still in its melted state to avoid pulling the track off the board. Generally, this leaves the pad relatively solder-free allowing you to easily pull the component off the board. If it does not easily come off, we recommend adding a little more solder and repeating. Make sure that you give the board time to cool down though as you can easily damage the PCB and the components on it by applying too much heat for too long. If you’re new to desoldering components we highly recommend you practice a little on old faulty boards from broken equipment. It’s a very handy skill to learn and a great way to start your component collection as many faulty boards still have functioning components like capacitors and inductors.

With the USB connector removed, you will see the four bare pads as shown here.

Note: Any brown excess flux residue from the desolder braid and the solder should be removed with methylated spirits or Isopropyl Alcohol to avoid causing damage to the PCB over time.

Circuit Board

We created a tiny circuit board to house the optocoupler and the 100Ω resistor for this project. It has a 3-pin female header on it that slides directly onto the male header of the PIR sensor. The wire to the camera and 2.5mm stereo plug is soldered directly to this board.

Once everything is soldered and connected, we recommend using liquid electrical tape to insulate the exposed connections. This is because the front and rear parts of the enclosure both have circuitry which could potentially come into contact and create a short-circuit. Since there isn’t any battery short circuit protection in this circuit, this could cause the battery to release a very large amount of energy quickly which could result in a fire.

If you don’t have liquid electrical tape, normal electrical tape or even a thin plastic barrier would be sufficient.

CONNECTING LEAD

The lead from the device to the camera was salvaged from an old set of mobile phone headphones. We added the coil to the lead as it is an easy way to prevent tangles when storing the device in your bag. We coiled it by wrapping the lead around a metal pipe, securing both ends with a clip, then used a hot air gun to heat the insulation. This causes the plastic/silicone insulation to soften and take the new shape. Once cooled, the lead keeps this shape. We fed the wire through the enclosure and tied a knot in the wire to work as a strain relief and prevent it from being pulled through.

3D PRINTED ENCLOSURE

We designed the enclosure in Fusion360 and will make the working files available so you can modify the design and make changes as you see fit.

We opted to use the GoPro mounting system as there are numerous official mounting options for these cameras, as well as many custom styles available on Thingiverse to suit different needs. If you want a custom mount, you can easily search Thingiverse for GoPro mount and you will find a huge amount of custom mounting options, saving you the time to design and model your own.

We wanted to have the device mount onto the camera itself, so before designing the hotshoe mount, we first did a search on Thingiverse and found that user Shantmeg had already designed and uploaded the exact part we wanted (https://www.thingiverse.com/thing:98700). Therefore, we didn’t need to model our own iteration and rather simply printed their design.

CASE FRONT

The front part of the enclosure was printed at 300-microns on our Flashforge Creator Pro using Flashforge branded translucent red PLA. The design needs support under the mounting bracket and the switch housing. In this configuration, the design takes around 1 hour to print.

Note: The hole for the lead to exit the enclosure has been made at 3mm. You may want to increase this diameter depending on the wire you use. You can do this using the working files we have made available or you can just import the STL file into Tinkercad and make the modification there.

CASE REAR

The rear of the case was printed at the same 300-micron layer height on our Flashforge Creator Pro and using the same Flashforge red PLA. In this configuration, it takes around 30 minutes to print and does not need any support material. The enclosure simply press fits together after all of the components have been installed.

HOT SHOE

The GoPro type hot shoe mount (Credit to Shantmeg on Thingiverse) has been designed with a small raft and supports included. Our print was made at 100-micron layer height on our Flashforge Creator Pro in the same translucent red Flashforge PLA. It took about 45 minutes to print in this orientation.

ASSEMBLY

The assembly of the device is fairly straightforward. You first want to secure the sensor so that the potentiometer faces the top of the device lined up with the holes in the enclosure. We have added standoff supports to the enclosure that will allow you to secure the sensor to the enclosure by using two M2 screws. However, if you don’t have screws, using hot glue or another adhesive will work too. Just be sure to not get any adhesive material on the potentiometers.

Insert the slide switch into the alcove designed into the enclosure, then secure it with an adhesive of your choice (We used hot glue). Be sure not to get any of the glue into the switch as it will prevent its actuation.

Glue the lithium battery charger module to the enclosure so that its USB connector sits into the cutout. After this, glue the boost module (with its USB socket removed) into the rear enclosure so the components are facing up. It is then just a matter of making all of the wired connections to the devices and the small perfboard. Make sure you solder the wires as quickly as possible to avoid melting the enclosure.

Note: You could also test fit everything before gluing them down and cut the wires to suit. You can then solder everything together and glue the final result into the enclosure.

Finally, plug in the Lithium battery.

TESTING

With the assembly complete, you can simply set the camera up on a tripod and use your chosen mounting system to mount the device in its desired location.

You will need your camera set to AI servo mode, which will force the camera to constantly check and maintain focus.

Note: On some Canon cameras, such as the 80D, the AI servo mode is only active when in live view mode. If your camera is not focusing correctly, double check by putting the camera into live view.

To get the best results from the autofocus, we found it was best to have as many AF points active as possible. This way, the camera can choose the best focus over the greatest area possible. Read your camera’s manual to find how to set this.

BATTERY LIFE

We made this project battery powered so there is no need to be tethered to a power supply. This does mean that the battery life needs to be considered, and should last at least as long as the camera it is attached to. The Canon 80D we were using for testing lasts for around 1.5 hours in live view mode (with screen brightness turned all the way down) so we figured this should be our benchmark. To calculate how long we can expect the battery to last we need to know the battery capacity and the amount of current the device draws.

To measure the current the device draws we removed the battery from the circuit. We set our bench power supply to 3.2V, which is the cutoff voltage of a lithium cell, and thus should mean the highest current draw needed to generate the 5V from the boost converter. We attached the ground of our power supply to the ground of the circuit and the positive to the positive probe of our Uni-T UT120C multimeter. We then connected the negative probe of the multimeter to the Vcc of our circuit.

We measured an average current of only 1.5mA, which jumped to 5mA when triggered.

The battery we used claims to have a capacity of 550mAh, so the battery life should be approximated using the following formula:

Battery Life = ( Battery Capacity (mAh) / Load Current (mA) ) x Other Factors*

Battery Life = ( 550 / 1.5 ) x 0.5 = 183 hours

*Other factors are included to account for factors such as efficiency, temperature variations and other such factors that are present but not easily calculated. We are dividing the results by half to err on side of caution.

This means at the very best case, with no triggering, the device will last for 183 hours (over 7 days), whereas in the worst case, the device will last for about:

Battery Life = ( 550 / 5 ) x 0.5 = 55h

This suggests that the battery will be more than sufficient to run this device for the few hours we are intending per charge and shows that the battery we chose could easily be replaced with a much smaller one.

A CR2032, for example, has a capacity of about 235mAh and thus, according to these calculations, it would last 80 hours in the best case. This would be more than sufficient for this type of device. However, its worth noting that the actual capacity is generally lower than the rated capacity and is detrimentally affected by things such as temperature and age.

PIR DETECTOR RANGE

These sensors have a working range of around 3–6 metres from the sensor with an angle between 110° and 70°. The sensitivity can be adjusted via one of the potentiometers while the other sets the duration of the high pulse after detection has occurred. Our testing was done with both potentiometers set to their lowest value, making it the most sensitive to changes and the duration very small.

We tested the device by setting the camera up on a tripod with AI servo/constant autofocus enabled with the motion trigger mounted to the hotshoe. We then walked in sideways from both the left and right of the setup, initially from 2m and repeating the test while increasing the distance by 1m.

Whilst we had no way to accurately measure the detection angle and distance the experiment was able to validate the claimed specs. At 6m from the sensor, walking parallel to the sensor, the detecting angle seems to be around 110°, whilst at 3m the angle is much smaller, presumably around 70°.

We also observed that the sensor often fails to detect movement when the object moving is directly in front of, and moving toward the sensor. In fact, the PIR sensor does NOT detect radial motion from any angle, as the animal needs to pass around in an arc to have their heat image detected via as many facits as possible to make the sensor see them.

Therefore, set the trap so the animal will pass across the front of the lens, not towards it.

PIRs as intruder detectors must also be placed diagonal to the intruders likely path, maybe vertically if you note the Fresnal pattern on a security PIR.

WHERE TO FROM HERE?

There are quite a few directions one could take this project to create some fairly unique and niche devices. Imagine, for example, rather than using an expensive DSLR, you used an inexpensive camera module and essentially make the project so inexpensive it would be disposable.

You could use something like the ESP32 CAM module in conjunction with this PIR module to work as a burglar alarm. When the PIR detects movement, it turns on the camera which starts recording, streaming or uploading to the cloud. When movement stops the camera shuts down conserving battery life.

The reality is, this sensor is so incredibly simple to use due to its very simple on or off output that it could be used in a huge number of projects where detecting the presence of a person or medium-sized animal is needed.